linux-kernel - RE: [PATCH v2] driver: uio: fix possible memory leak and use-after-free in __uio_register

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <4F88C5DDA1E80143B232E89585ACE27D0258ECAE@DGGEMM528-MBX.china.huawei.com>
Date:   Tue, 8 Jan 2019 02:51:01 +0000
From:   "liujian (CE)" <liujian56@...wei.com>
To:     Greg KH <gregkh@...uxfoundation.org>
CC:     "michal.simek@...inx.com" <michal.simek@...inx.com>,
        "hamish.martin@...iedtelesis.co.nz" 
        <hamish.martin@...iedtelesis.co.nz>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: RE: [PATCH v2] driver: uio: fix possible memory leak and
 use-after-free in __uio_register_device


> -----Original Message-----
> From: Greg KH [mailto:gregkh@...uxfoundation.org]
> Sent: Tuesday, January 08, 2019 12:13 AM
> To: liujian (CE) <liujian56@...wei.com>
> Cc: michal.simek@...inx.com; hamish.martin@...iedtelesis.co.nz;
> linux-kernel@...r.kernel.org
> Subject: Re: [PATCH v2] driver: uio: fix possible memory leak and use-after-free
> in __uio_register_device
> 
> On Fri, Jan 04, 2019 at 10:19:08PM +0800, liujian wrote:
> > 'idev' is malloced in __uio_register_device() and leak free it before
> > leaving from the uio_get_minor() error handing case, it will cause
> > memory leak.
> >
> > Also, in uio_dev_add_attributes() error handing case, idev is used
> > after device_unregister(), in which 'idev' has been released, touch
> > idev cause use-after-free.
> >
> > Fixes: a93e7b331568 ("uio: Prevent device destruction while fds are
> > open")
> > Fixes: e6789cd3dfb5 ("uio: Simplify uio error path by using devres
> > functions")
> > Signed-off-by: liujian <liujian56@...wei.com>
> > Reviewed-by: Hamish Martin <hamish.martin@...iedtelesis.co.nz>
> > ---
> > v1->v2:
> > change git log and fix code
> >
> >  drivers/uio/uio.c | 10 +++++++---
> >  1 file changed, 7 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/uio/uio.c b/drivers/uio/uio.c index
> > 1313422..be2a943 100644
> > --- a/drivers/uio/uio.c
> > +++ b/drivers/uio/uio.c
> > @@ -940,9 +940,12 @@ int __uio_register_device(struct module *owner,
> >  	atomic_set(&idev->event, 0);
> >
> >  	ret = uio_get_minor(idev);
> > -	if (ret)
> > +	if (ret) {
> > +		kfree(idev);
> >  		return ret;
> > +	}
> >
> > +	device_initialize(&idev->dev);
> >  	idev->dev.devt = MKDEV(uio_major, idev->minor);
> >  	idev->dev.class = &uio_class;
> >  	idev->dev.parent = parent;
> > @@ -953,7 +956,7 @@ int __uio_register_device(struct module *owner,
> >  	if (ret)
> >  		goto err_device_create;
> >
> > -	ret = device_register(&idev->dev);
> > +	ret = device_add(&idev->dev);
> >  	if (ret)
> >  		goto err_device_create;
> >
> > @@ -985,9 +988,10 @@ int __uio_register_device(struct module *owner,
> >  err_request_irq:
> >  	uio_dev_del_attributes(idev);
> >  err_uio_dev_add_attributes:
> > -	device_unregister(&idev->dev);
> > +	device_del(&idev->dev);
> >  err_device_create:
> >  	uio_free_minor(idev);
> > +	put_device(&idev->dev);
> 
> device_del() and then put_device()?  I don't think that's a correct error
> cleanup path do you?
> 
In err_uio_dev_add_attributes error handling case, device_unregister()  will call put_device(), and idev will be released in put_device()
Then in err_device_create error handling case,  uio_free_minor() will trigger use-after-free

Now, divide device_register() into device_initialize() and device_add(), and divide device_unregister() into device_del() and put_device(). 
Call uio_free_minor() and then call put_device() to avoid use-after-free

> Please fix one thing at a time here also, this should be a a patch series, right?
> 
Yes, I will do, thanks~

Best Regards,
Liujian

> thanks,
> 
> greg k-h